Longitudinal data on factors predictive of the devlopment of hypertension in middle age are sparse. Taking advantage of large numbers of multiphasic health check-ups given to Kaiser-Permanente subscribers we propose to (1) test longitudinally some established hypotheses concerning obesity, alcohol use, family history and salt use and, more important, (2) explore over 800 variables in search of presently unrecognized predictors of hypertension. A pilot study indicates that we can identify an estimated 1,343 persons, age 30-49 at entry, with documented normotension progressing to documented essential hypertension. For each such case we will select one age-sex-race-matched control with persistent normotension. Both characteristics at baseline and changes in characteristics will be examined as possible predictors of hypertension. To minimize the likelihood of chance positive findings we will split the sample in half and determine whether positive findings in the first half are also present in the second. Since hypertension represents such an important public health problem and its treatment is not entirely benign, it is desirable to attempt to find new avenues for prevention where a good opportunity to do so exists.